This New York skyscraper had a 1-in-16 chance of collapse.

[Alan Ambrose]

I think Gus has mentioned this before, but there's a new book on LeMessurier:

 

This New York skyscraper had a 1-in-16 chance of collapse. Only one man knew | CNN

[SteamyTea]

2 hours ago, Alan Ambrose said:

a new book on LeMessurier

I posted up a YouTube video about it.

Is fascinating.

 

 

[Gus Potter]

12 minutes ago, SteamyTea said:

Is fascinating.

It is. When I was at uni I had a lecturer that was about ten years older than my self, I went to uni at 40 he was about 50 and had a long an varied experience first as a Civil Engineer (like @saveasteading) and then got into structures. He used this as a teaching example about how you need as an Engineer to "sense check things". A good Engineer uses a computer, maybe complex analysis (say FE)  to refine the design, seek economy.. but we always use the model as a tool. We never take computer output and sign off unless we have sense checked it. That is dangerous. 

 

Another example is the Hyatt Hotel collapse. This came about because of; a simple steel detailing error, if the detailer/ engineer had put a big red note on the drawing.. don't (expletive deleted) about! , the contractor did not change things on site ( explain the buildability) all coupled with a loss of engineering oversight and site inspection.

 

https://en.wikipedia.org/wiki/Hyatt_Regency_walkway_collapse

 

Now in terms of self building.. always think before you let your contractor go off and do their own thing! Don't let them sway you.. we have always done it this way and we know best! The modern self build is becoming a lean design animal.. to make it worth while (you make some money) designers are pushing the boundaries and with that comes a need for builders to get with the programme!

 

 

[saveasteading]

9 hours ago, Gus Potter said:

we have always done it this way and we know best!

This is so common. They don't know how little they know.

@Gus Potter fan you recall what safety factor we add for workmanship? But it still  assumes some competence and good intentions.

[Gus Potter]

On 20/08/2025 at 07:52, saveasteading said:

you recall what safety factor we add for workmanship? But it still  assumes some competence and good intentions.

It depends.. but for all self builders out there. Often in my sometime lengthy submission to BH (this is kind of one of them)  I refer to the Bristish Standards as they are often easier to understand and still applicable particularly to self building. But in this case to get a handle on things we are better to look to the Euro codes as they expose more of this factor of safety malarky cf some of the Bristish standards which work on permissible stress, not that many mind, but enough.. all the safety factors get lumped together. 

 

Let's take a steel portal frame first. This could be an agricultural barn conversion. Most design codes recognise that many are put up out of plumb, the steels are not straight due to the hot rolling process. If you have seen steel erectors work, they have a sledge hammer in the van, they use drift pins, crow bars, the tele handler (when no one is looking) and straps to pull things into place to get the connection holes to line up for example. When it comes to levelling base plates, you maybe need to go and have a tea break if you are a Client / Engineer.

 

To help us get round this we assume that the frame is erected out of plumb by 5.0 mm for every meter height. 5.0mm  / 1000mm is 0.5% gradient. This means that on a 6.0m high steel portal the roof could be sitting off centre from the base plate by 6.0 x 5.0 = 30mm. We then calculate this eccentric load and apply as SE's in our design,  this is what we call a notional horizontal load, which is basically a saftey factor on workmanship and manufacturing. You'll also find this figure in the British Constructional Steel Association guide to Steel erection which describes an acceptable  level of workmanship that can be reasonably expected on site. 

 

Ok turning now to timber design. Typical thing you see on BH is folk using joist hangers and angle brackets. If you take say a Simpson Strong ABR 9020 bracket. Here they declare the Charachteristic strength which then gets divided by a material factor of commonly 1.3 to give the design strength, sometimes shown as the design load. Don't mix this up with a safe working load as the loads also need to be factored UP first. There is no workmanship safety factor here. The workmanship is controlled by other bits of the design codes where we make sure the nails have adequate edge / end distances and so on. If the chippie / joiner does not follow the drawings or has not been properly trained there is no fall back position. 

 

Masonry.. bricks and blocks.  Here again we rely on, as does many other aspects of the build porcess,  on the British Standard 8000 series which relate to workmanship and tolerances. But qualitively we could look at the BS 5628 part 1 for a bit of ball park guidance. Below is a screenshot of the masonry table. 

 

 

Normally you see SE's using the bottom right saftey factor of 3.5 as the factor by which we reduce the strength by. So say you are using a 7.0 N dense concrete block in an class (iii) M4 mortar. Characteristic strength is 6.4 N/mm sq. Take the 6.4 / 3.5 = 1.83 N/mm^2 or 6.4 / 2.8 (bottom left figure) = 2.29  N/mm^2. So in terms of construction control you get 2.29 / 1.83 = about a 25 increase in performance. But on a self build you just won't spec this as the reality is that the self builder will never be able to execute this, and to find a brickie that can is whishful thinking, then you have the additional labour cost. 

 

 

In the round then we need to establish a reasonable quality of workmanship. Here we may look to the NHBC guidelines and standards for tolerances on the masonry. 

 

In summary your workmanship needs to be to the design codes or manufacturer's recommendations. The manufacturer's need to sell stuff and it's fair for them to assume you can do the work to a reasonable recognised standard.

 

 

 

 

 

 

 

[Iceverge]

20 minutes ago, Gus Potter said:

Let's take a steel portal frame first. This could be an agricultural barn conversion. Most design codes recognise that many are put up out of plumb, the steels are not straight due to the hot rolling process. If you have seen steel erectors work, they have a sledge hammer in the van, they use drift pins, crow bars, the tele handler (when no one is looking) and straps to pull things into place to get the connection holes to line up for example. When it comes to levelling base plates, you maybe need to go and have a tea break if you are a Client / Engineer.

 

To help us get round this we assume that the frame is erected out of plumb by 5.0 mm for every meter height. 5.0mm  / 1000mm is 0.5% gradient. This means that on a 6.0m high steel portal the roof could be sitting off centre from the base plate by 6.0 x 5.0 = 30mm. We then calculate this eccentric load and apply as SE's in our design,  this is what we call a notional horizontal load, which is basically a saftey factor on workmanship and manufacturing. You'll also find this figure in the British Constructional Steel Association guide to Steel erection which describes an acceptable  level of workmanship that can be reasonably expected on site. 

 

I'm self building a small cattle shed at them moment. 

 

I laughed out loud at this. 5mm per m is about as accurate as I could manage. I then got bored of using ratchet straps to straighten my stanctions so I gave them a shove of the backhoe. 

 

 

Nothing lines up properly in the real world and I've been using clamps and threaded bar to pull and cajole every bracket into place. 

 

I wouldn't like to do it on a skyscraper but I dare say it might happen more than I realise. 

 

 

 

[Gus Potter]

On 19/08/2025 at 21:25, SteamyTea said:

Is fascinating.

Yes I think I mentioned this a few years back. It was the quatering wind that blew the gaff.

 

It was from memory a student at uni that was looking for a dissertation project and that is where the "discovery" started. 

 

There are a few on BH that are involved or have been in education. One great example here is Jeremy Harris. He wrote some design codes, has a huge wealth of knowledge. From time to time I discuss with other Engineers / designers what we call loss of grand father experience / loss of tacit knowledge. You get to call yourself a designer when you have enough experience to appreciate that what you do is an art underpinned by your theoretical knowledge. 

 

One theme that crops up a lot is that the latest generation, given the information over load that young folk experience means that they just don't have time to look back two generations and learn from some of the horrific mistakes. Tacoma narrows brige is a good video!

 

The other big problem is that young Architect's, SE's and other construction professionals are under a lot of commercial pressure which leads them in to more specialist areas of design.. they just don't have time to learn more about the general design craft and that leads to mistakes.

 

In summary you only need to be a self builder trying to get your TF manufacturer to talk to say your Achitect or SE to see the results of this. 

 

 

 

[Gus Potter]

13 minutes ago, Iceverge said:

I laughed out loud at this. 5mm per m is about as accurate as I could manage. I then got bored of using ratchet straps to straighten my stanctions so I gave them a shove of the backhoe. 

Well that is my good turn for the day, make folk laugh! Glad to see you used the backhoe as much safer than trying to yank it with a telehandler? 

 

To be serious, looks like you have done the job getting that to 5.0mm per m, well done!

 

What's going on with the top of your column to rafters connections? I don't know where you shop for your clothes but these look a bit like a "summer blouse". 

 

Are you going to leave one side open. Any ideas on wind bracing as yet? 

 

 

[Iceverge]

 

Lots of this going on. I'm working it out as I go. 

 

 

 

I'm going to add some to the low side too. 

 

I'm planning on pouring mass concrete walls to about 1800mm high along the long low side and the short side nearest to the camera on the lower picture. 

 

Above these I will most lightly add Yorkshire boarding for ventilation. 

 

The long side near the digger I think I'll pour a low wall, maybe 400mm.and the other side will be open. 

 

I'm very much working on "does it look right" and "give it a push and if dissatisfied add more bracing"  methodology.  It's more substantial than most sheds I've seen locally but of course that's no guarantee. 

 

You mentioned earlier about barn conversions. I think this should be a mandatory lesson to anyone undertaking one on the fly by night farmer engineering that occurs.

 

[Gus Potter]

44 minutes ago, Iceverge said:

I'm very much working on "does it look right"

Ok it's a shed, call it agricultural. I used to do the SE calcs that underpinned the software that was used to design these sorts of things. So technically I've had a hand in, as an SE, in thousands of these types of frame, all over the world and not just in the UK. So I know a bit more than many SE's about this. 

 

But let's be pragmatic and look at the risk if it falls down. First thing is we don't wan't anyone to get hurt. To reduce the risk we want make sure that folk are not near it or in it when it snows a lot or it is windy. In agricultural design codes we call this the duration an occupancy rating.  If it was a chicken shed then you would only be in it say 2 hours a day and there maybe only two folk at the most at any time. 

 

Now the other thing we look at is how close to the boundary your shed is. If the roof flies of your shed and hits a house or a car driving along the road that is not good. In the agricutural design codes there is a bit that deals with how close to the road it is. 

 

So once we have dealt with the risk you may or may not want to get some insurance. Up to you but some places like yours are next to care homes and schools.. catch my drift. 

 

Now once we have satisfied ourselves that we have not done something that is downright dangerous we maybe want to look as how the building moves about. Is is going to move differentially at roof level so much that within a few years the roof cladding fixing start to oval and leak.. then we have wasted out money or if it is just for that we can we live with a bit of a leaky roof? 

 

1 hour ago, Iceverge said:

I'm planning on pouring mass concrete walls to about 1800mm high along the long low side and the short side nearest to the camera on the lower picture.

Don't do this unless you design them to be free standing.. it's bloody dangerous other wise.  Or you can get an SE to sort this out for you properly. 

[saveasteading]

On 19/08/2025 at 21:56, Gus Potter said:

We never take computer output and sign off unless we have sense checked it. 

 

To my mind the building in question fails this on simple inspection.

What if one of those columns moves...the building is inherently precarious. 

We may move to discussing disproportionate collapse , and Limit State Design.

 

And to rely on damping is playing games on such a huge building. Of course we must experiment, but test it first and overdesign if anything.

Interesting to read that the steel fabricators made changes without consultation.

 

9 hours ago, Gus Potter said:

to get the connection holes to line up for example. When it comes to levelling base plates, you maybe need to go and have a tea break if you are a Client / Enginee

On entering the portal frame construction world,  i was shocked to see portal frame buildings being erected on foundation bolts that moved to wherever the steel bases were.  These are in cones and are grouted up after. The tolerances were huge:  30mm typically?

So when I had control I had all bolts cast  in solid, in the right position to a few mm, and heights likewise.

Concretors  hated having to get it right, so I became an early adopter of epoxy fixing: drilling into solid concrete.

The steel went up beautifully thereafter.

Perhaps the ridges were 30mm out as  @Gus Pottersuggests, but that is trivial in a 30m span.

A squarely lined up frame goes up much better too and there a lot of bolt holes to line up.

[saveasteading]

6 hours ago, Iceverge said:

planning on pouring mass concrete walls to about 1800mm high

 I've designed and site worked on loads of such walls, for bridges, reservoirs etc. They are serious engineering with the shuttering and supports costing more than the concrete. The sideways load from wet concrete is huge. 

You will also need 2 layers of reinforcement.

And tying them to your superstructure, which holds up the other?

Blockwork is more appealing. 

Is there a reason why you fancy or need concrete walls? If it's a grain store, the grain acts like a fluid, trying to knock your wall over.

[Iceverge]

Thanks for the input folks. 

 

It's a couple of hundred metres from the nearest houses / public road and human occupancy will be for strawing. I would say 15-30 mins per day. 

 

7 hours ago, Gus Potter said:

Don't do this unless you design them to be free standing.. it's bloody dangerous other wise.  Or you can get an SE to sort this out for you properly. 

 

I had planned on tying it into the floor with A393 mesh, maybe in an "L shape" not sure if this counts as free standing. It will also tounge into the groove in the stanctions ( old railway track by the way from the Cork Dublin line I was told) 

 

2 hours ago, saveasteading said:

my mind the building in question fails this on simple inspection.

What if one of those columns moves...the building is inherently precarious. 

We may move to discussing disproportionate collapse , and Limit State Design.

 

Yes it's still quite ricketey. I'm adding more bracing again today. I'll keep going until I'm happy. I didn't mention but the stanctions are set in concrete to approx 700-1000mm deep so it's not quite so bad as it would be if they were just on base plates. 

 

1 hour ago, saveasteading said:

The sideways load from wet concrete is huge. 

 

Don't I know it! Having seen a shutter let go is no fun. 

 

1 hour ago, saveasteading said:

You will also need 2 layers of reinforcement.

 

Noted but there may not be enough width. I had planned on a 160mm wall on the long side with a single mesh biased towards the outside . Maybe a 50/110mm split. I will have a 200mm wall to the short side. Again maybe a 60/140mm split as all the pushing force on the wall will be from the inside.

 

1 hour ago, saveasteading said:

And tying them to your superstructure, which holds up the other?

 

Both in my head. The concrete walls between the stanctions will resist racking under compression. The flanges of the stanctions will stop the concrete walls pushing in and out. 

 

1 hour ago, saveasteading said:

Blockwork is more appealing. 

 

I'm surprised you say this. I have seen block walls fail under very little pressure from animals (think a dozen 600kg bullocks taking fright) . Equally I will need to use the backhoe to clean the shed out. It would simple drive straight through a block wall if I got it wrong. 

 

If you look to the right of the picture with the backhoe there is a stub wall approx 450mm high and 150mm wide with some 252 or 393 mesh in it under the feed barriers.  I push silage against it with the backhoe and thus far it's held up well. 

 

Thanks again for the input.

 

[Iceverge]

 

That's a better pic of the said stub wall. 

[saveasteading]

5 hours ago, Iceverge said:

all the pushing force on the wall will be from the inside.

Your logistics is back to front. The reinforcement works best in tension, so centrally or inside.

The reason I suggested block is for construction and avoiding shuttering. Try cavity 2 x 100. Infill slowly with concrete if you want.

 

5 hours ago, Iceverge said:

A393 mesh, maybe in an "L shape" not sure if this counts 

Sorry.  I see you're playing.

I will duck out now.

Just be sure never to let children play in there, and knock it down before passing on the property.

 

[Alan Ambrose]

@saveasteading >>> So when I had control I had all bolts cast  in solid, in the right position to a few mm, and heights likewise.

 

Any more info on how you got them so accurate?

[SteamyTea]

44 minutes ago, Alan Ambrose said:

Any more info on how you got them so accurate

Years not drinking at university helps.

And a good tape measure.

[Alan Ambrose]

>>> Years not drinking at university helps.

 

It’s a bit late to be telling me that now.

[Iceverge]

6 hours ago, saveasteading said:

Your logistics is back to front. The reinforcement works best in tension, so centrally or inside.

 

I'm not sure I understand. 

 

The vertical stanctions I have in my head as orange. With the main load coming from the inside (cattle/digger) I was biasing the reinforcement (red) towards the outside as this was where I saw as the most tension . 

 

 

6 hours ago, saveasteading said:

Sorry.  I see you're playing.

 

Not at all. What point did I lose you? 

[saveasteading]

5 hours ago, Alan Ambrose said:

Any more info on how you got them so accurate

 

4 hours ago, SteamyTea said:

good tape measure.

Easiest with threaded rod and epoxy.

Presumably there is a datum for each wall that is simply measured off. Alternatively use strings or fancier laser type stuff. The bolts are measured as a cluster relative to that.*

For height use steel shims set to the height required. I preferred to set sacrificial nuts on the bolts, lowered to the level, thus typically four bolts support the column until it is grouted underneath.

Happy to expand if you explain which bits in particular.

Perhaps on a new thread.

 

* although skilled in theodolites etc, I prefer lines and tape measures. It also means that the building usually has an edge beam by that stage, so it's easy access and to mark out. If the perimeter is built then it's even simpler.

Edited 14 hours ago by saveasteading

[saveasteading]

4 hours ago, Iceverge said:

What point did I lose you? 

I said I'm out so should maintain that.

Briely then, in the retaining wall picture the base is fixed so doesn't rotate. The top deflects outward, so the inner face is in tension.. However  the pc  retaining wall does not apply acc to the orangy sketch

All that load goes onto the skinny columns which fall over.

 

[Nickfromwales]

If you look back at @Jeremy Harris build, his retaining wall would fend off WWIII in comparison. 

[SteamyTea]

3 hours ago, Nickfromwales said:

retaining wall

The gabion one, or the one along the road?

[Nickfromwales]

Just now, SteamyTea said:

The gabion one, or the one along the road?

The cast one. 

[SteamyTea]

1 minute ago, Nickfromwales said:

The cast one. 

Think you mean the gabions behind his house, the one his neighbour wanted to have a fence on.

Was not cast, or rendered when I first went there, may be now, did not look last visit.

Looked at his new Sunamp.